The Scottish Government

« Previous | Contents | Next »

Listen

Technical Appendix A4 Simulation results: Costly Requirements on Households to Reduce Energy Use

A4.1 Introduction

In this section we attempt to model the impacts of policies directed at household energy use. Currently, it is not possible to directly simulate issues relating to household energy use using the AMOSENVI framework. ¹⁸ Instead, here we attempt to model the knock on effects of such a policy (resulting from the impact on household income). In this Appendix, we describe the background and simulation strategy we use to model such a scenario using the AMOSENVI model and present the results from our simulations. Some sensitivity analysis is conducted over key parameters, including the assumed structure of the Scottish labour market in order to highlight the importance of these variables for our central results. As with other policies modelled, these simulations assume that the policy is directly introduced in Scotland, but not the rest of the UK. As such, these simulations can be considered as showing the impact of a differential policy introduced in Scotland, which goes beyond that of any policies introduced at the level of the UK as a whole. The impacts of such policies are those over and above the impacts of any UK-wide policies introduced.

This appendix is structured as follows. In Section A4.2 we set out our simulation strategy, and then in Section A4.3 we report the economic, environmental and energy results for the range of shocks to household income considered. In sensitivity analysis in Section A4.4, we vary the size of the decrease in household income, the migration closure and the labour market specification of the AMOSENVI model.

A4.2 Simulation strategy

In AMOSENVI we can model the labour market and system-wide consequences of the introduction of policies that serve to reduce household income. Such an approach is consistent with a mandatory requirement for households to purchase costly technologies that may reduce their energy use. There will be system-wide labour market consequences of the implied reduction in household income. The reduction in household income will lead to workers bargaining for an increased nominal wage, which will in turn reduce the competitiveness of Scottish economic activity. We would also expect there to be migration effects as, in AMOSENVI, net migration to Scotland is driven by real wage and unemployment rate differentials between Scotland and the rest of the UK. A lower real (take-home) wage may induce migration from Scotland.

The labour market impact of such a reduction would operate in a similar way to the imposition of an appropriately calibrated increase in income tax with no recycling of the additional revenue back into the economy ¹⁹. This is the way that we model the impact in the simulations presented here. We model the impact of a 1% decline in household income. This is achieved through raising the overall share of wages paid to income tax by 1%. This requires a 6.3% increase in the rate of income tax in the base year data ²⁰.

A4.3 Central scenario results

Central aggregate and sectoral results

As with previous simulations, results should be interpreted as being variations away from what would have happened to economic activity and environmental impacts but for the policy that reduced household income. Table A4.1 shows the aggregate results for economic, energy and environmental from such a policy in the long-run. The long-run here is a conceptual time period over which labour and capital stocks have fully adjusted to new equilibrium levels. In AMOSENVI with migration possible, this is consistent with a time period over which the real wage and unemployment rate have been restored to their initial equilibrium values, and the capital rental rate is equalised across all sectors in the economy.

Table A4.1: Short- and long-run impacts on aggregate economic, energy and environmental indicators under a 1% decrease in household income, bargaining labour market, % changes from base year

In line with our expectations, we observe that the initial decrease in household income leads to a 0.35% fall in the real wage in the short run ( i.e. while capital and labour stocks are fixed). (Private) consumption is down by 0.80%, and overall GDP is lower by 0.23%. Under a bargaining labour market specification with migration from and to Scotland possible, as is used here, we would expect that this would lead to outmigration. While population is fixed in the short-run, over the long run outmigration should act to restore the real wage differential between Scotland and the rest of the UK.

By the long-run, GDP is 1.63% lower, and while real wages have risen back to their pre-shock levels, nominal wages are 1.33% A4.1 higher and the CPI is higher. This has a damaging impact on employment and exports. The sectoral pattern of changes in output and employment is shown in Figure A4.1. Sectors that are labour intensive and export intensive suffer particularly badly, as (before tax) wages are higher and higher prices damage the competitiveness of output.

Figure A4.1: Long-run impact on sectoral output and employment, % changes from baseyear

The environmental consequences of this policy are lower emissions in both the short and long-run, and the decreases in CO2 emissions are greater than the falls in GDP. The CO2 intensity of production thus decreases. The time path of the changes in GDP, CO2 emissions and the sustainable prosperity measure are shown in Figure A4.2. The simulation is run over 150 periods in order that a long-run equilibrium is reached, although most of the adjustment to the long-run has occurred by period 120. The CO2 intensity of production falls immediately and is lower again in the long-run, but does not decrease monotonically before reaching its long-run equilibrium.

Figure A4.2: GDP, CO ₂ emissions and CO ₂ intensity of production following a 1% decrease in household income, bargaining labour market, % changes from base

Energy demand (both electrical and non-electrical energy demands) is lower in the short- and long-run compared to the base year, with greater reductions in the long-run. The GDP/energy indicators show positive movements in sustainability, i.e. increasing GDP per unit of energy. Positive changes in this variable show greater economic output per unit of energy, and, despite total GDP being lower - both electrical and non-electrical energy use shows a greater decline. The profile of adjustment between the short-run and long-run equilibrium path for electrical and non-electrical energy demands - along with GDP and the GDP/energy indicators - is shown in Figure A4.3.

Figure A4.3: GDP, electrical energy and non-electrical energy demands following a 1% decrease in household income, bargaining labour market, % changes from base

A4.4 Sensitivity analysis

Sensitivity to size of the decrease in household income

Table A4.2 shows the aggregate long-run economic, energy and environmental results for alternative decreases in household income, under a regional bargaining labour market specification. As can be seen, when the decrease in household income is changed, the results across all variables respond in an approximately proportional way - i.e. a 0.5% reduction in household income has results roughly approximate to one quarter the effect of a 2% reduction in household income, and one half those of a 1% reduction in household income. Across all values of the shocks, the positive movements in the CO ₂ intensity of output and the GDP/energy indicators remain, although GDP continues to be lower than the base year under all scenarios.

Table A4.2: Percentage changes in long-run for aggregate economic, energy and environmental indicators under variations in the decrease in household income, bargaining labour market

Sensitivity to migration configuration of model

As mentioned above in the discussion of the 1% reduction in household income, an decrease in household income has the effect of lowering real (take-home) wages. In AMOSENVI, net migration to Scotland is a function of the real wage and unemployment differential between Scotland and the rest of the UK. Lower real wages in our simulations reported to date, will cause outmigration to occur as workers seek higher wages in the rest of the UK. This migration will reduce the size of the labour force in Scotland, and would be expected to exacerbate any decline in economic activity in Scotland that comes directly as a result of the lowering of household incomes. To that extent, it would be useful to consider a case in which such economic migration between Scotland and the rest of the UK is not permitted. Table A4.3 shows the aggregate results for a 1% decrease in household income in the cases where migration is and is not permitted.

Table A4.3: Percentage changes in long-run for aggregate economic, energy and environmental indicators under 1% decrease in household income with and without migration, bargaining labour market, % change from base

As expected, there are significant differences between the results with and without migration for a 1% decrease in household income. The primary cause of this is the lack of any substantial decrease in population and employment in the long-run. No migration means that real wages are lower in the long-run, with a smaller increase in nominal wages, and so a less dampened impact on economic activity. While, relative to the base year, employment is still lower in the long-run under the no-migration case, the reduction (-0.67%) is significantly less than that seen where migration is permitted (-1.66%). There is a much less pronounced decrease in exports and investment as well, supporting a relatively higher level of economic activity. The CO2 intensity of output improves, as in the central results presented earlier, as CO2 generation falls by more than the decline in GDP. Electrical and non-electrical energy demands again decrease, but the increases in activity per unit of energy are smaller than where migration is permitted.

Sensitivity to labour market specification of the model

All of the simulations reported above assume that the regional labour market can be characterised by wage bargaining specification, where regional wages are related to the tightness of the regional labour market. In this subsection, we present some results from alternative treatments of regional wage setting. Two extremes of the regional labour market are explored: exogenous (fixed) labour supply, which would be consistent with a perfectly inelastic labour supply curve; and a fixed real wage specification, implying a perfectly elastic labour supply curve, as might be associated with an Input-Output specification in which there are no labour supply constraints in the regional economy. As in the previous sub-section, we also explore the implications under each of these specifications when migration is and is not permitted.

Table A4.4 shows the aggregate economic, energy and environmental impacts by the long-run under the three labour market scenarios, and for two migration scenarios considered in this section for each labour market scenario.

Table A4.4: Percentage changes in long-run for aggregate economic, energy and environmental indicators under 1% decrease in household income with and without migration possibilities for bargaining, exogenous labour supply case and fixed real wage specification of the regional labour market, % changes from base year

Looking first at the Bargaining labour market specification, but comparing the case with Migration on, against the Migration off case, we see there are substantial differences in the long-run results. With Migration off, the long-run is found when capital rental rates are equalised across sectors, but we no longer assume that migration acts to restore real wages to their initial equilibrium levels. Real wages are lower under the Migration off case in the long-run and there is a smaller reduction in employment than the Migration on case. We also observe an increase in the unemployment rate in the long-run when Migration is not permitted. GDP falls by less in the Migration off case than in the Migration on case, protected by the decline in the real wage rate (and smaller increase in nominal wages) acting to partially offset the decline in employment and economic activity.

In both Bargaining Migration on and Migration off scenarios, the reduction in CO ₂ generation is greater than the reduction in GDP, improving the CO ₂ intensity of output indicator. Energy demand falls in both cases, less in the Migration off case, as would be expected by the relatively greater economic activity in this scenario. Both of the GDP per unit of energy indicators increase, indicating an improvement in the sustainability of energy use.

With a fixed real wage and with migration on, the long-run changes in economic activity, energy use and environmental impact will be the same as that under a bargaining labour market in which migration is also possible. In the long-run of the Bargaining case with Migration, the imposition of the 1% decrease in household income results in out-migration and a recovery of the real wage to initial equilibrium level. In the fixed real wage, the wage rate is fixed over all time periods. The results may be identical in the long-run, but the time path adjustments are quite different. Figure A4.4 shows the changes in GDP under all six of these labour market and migration specifications.

Figure A4.4: GDP changes over time under bargaining, exogenous labour market and fixed real wage labour market specifications, with migration on and off, % changes from base

« Previous | Contents | Next »